Issue 155 New Unit Test for Field

[LinhWuerzburger]

added two tests that control the data pointer after a change in the data field

[LinhWuerzburger]

we may consider removing the Field::swap method
as it is currently implemented it can cause problems for the GPU implementation.

For example:
Field a with pointer 123 to its field and
Field b with pointer 345 to its field
both send their field to the GPU through data_pointer_a = 123, data_pointer_b = 345
GPU stores 123 as well 345
Field::swap(a, b)
now it looks like this:
a with pointer 345
b with pointer 123
data_pointer_a = 123
data_pointer_b = 345

if you know forget to swap data_pointer_a and data_pointer_b as well it is doomed to fail. But if the developer has to use swap it is clear what happens. of course, he is free to look into the Field::swap method but I think that is an easy source for an error when you are not fully aware of what happens in the Field::swap method. Especially because you are calling the function with the Field objects and not the data pointer.
At least a warning is needed as part of the function's documentation, in case the function is not removed.

[LinhWuerzburger]

don't we need a present clause here?

[LinhWuerzburger]

same matter in lines 90 and 73

[LinhWuerzburger]

independent is also important to tell the compiler that the loop can be freely parallelise